redis集群调试1

jiaxp

　　Redis-cluster
　　前面redis集群搭建测试完了，接着开始调试：
　　1. Redis-cluster容错，master宕机以后slave选举升级为master；

　　7000master没出故障之前：
　　7003slave->7000 master
　　127.0.0.1:7001>cluster nodes
　　c500c301825ceeed98b7bb3ef4f48c52d42fe72d127.0.0.1:7003 slave b62852c2ced6ba78e208b5c1ebf391e14388880c 0 1478763693463 10 connected
　　6d04c7a1efa63ca773114444591d04f6d340b07d127.0.0.1:7001 myself,master - 0 0 2 connected 5461-10922
　　b62852c2ced6ba78e208b5c1ebf391e14388880c 127.0.0.1:7000 master - 01478763693968 10 connected 0-5460
　　79d8a5dc7e6140540c03fe0664df6a8150d0888f127.0.0.1:7004 slave 6d04c7a1efa63ca773114444591d04f6d340b07d 0 1478763692453 5connected
　　642a60187eea628a3be7fa1bb419e898c6b4f36a127.0.0.1:7002 master - 0 1478763692453 3 connected 10923-16383
　　bf2e5104f98e46132a34dd7b25655e41455757b5127.0.0.1:7005 slave 642a60187eea628a3be7fa1bb419e898c6b4f36a 0 1478763692453 6connected
　　故意宕掉7000端口
　　日志：
　　32138:M10 Nov 15:39:24.781 * FAIL message received from6d04c7a1efa63ca773114444591d04f6d340b07d about79d8a5dc7e6140540c03fe0664df6a8150d0888f
　　32138:M10 Nov 15:39:24.782 * FAIL message received from6d04c7a1efa63ca773114444591d04f6d340b07d aboutc500c301825ceeed98b7bb3ef4f48c52d42fe72d
　　32138:M10 Nov 15:39:24.867 * Marking node bf2e5104f98e46132a34dd7b25655e41455757b5 asfailing (quorum reached).
　　32138:M 10Nov 15:39:27.532 * Clear FAIL state for nodec500c301825ceeed98b7bb3ef4f48c52d42fe72d: slave is reachable again.
　　32138:M10 Nov 15:39:28.566 * Slave 127.0.0.1:7003 asks for synchronization
　　32138:M10 Nov 15:39:28.566 * Full resync requested by slave 127.0.0.1:7003
　　32138:M10 Nov 15:39:28.566 * Starting BGSAVE for SYNC with target: disk
　　32138:M10 Nov 15:39:28.567 * Background saving started by pid 32154
　　32154:C10 Nov 15:39:28.609 * DB saved on disk
　　32154:C10 Nov 15:39:28.610 * RDB: 4 MB of memory used by copy-on-write
　　32138:M10 Nov 15:39:28.666 * Background saving terminated with success
　　32138:M10 Nov 15:39:28.667 * Synchronization with slave 127.0.0.1:7003 succeeded
　　32138:M10 Nov 15:39:31.612 * Clear FAIL state for node 79d8a5dc7e6140540c03fe0664df6a8150d0888f:slave is reachable again.
　　32138:M10 Nov 15:39:36.121 * Clear FAIL state for nodebf2e5104f98e46132a34dd7b25655e41455757b5: slave is reachable again.
　　32138:signal-handler(1478764594) Received SIGTERM scheduling shutdown...
　　127.0.0.1:7001>cluster nodes
　　c500c301825ceeed98b7bb3ef4f48c52d42fe72d 127.0.0.1:7003 master - 01478764604059 11 connected 0-5460
　　6d04c7a1efa63ca773114444591d04f6d340b07d127.0.0.1:7001 myself,master - 0 0 2 connected 5461-10922
　　b62852c2ced6ba78e208b5c1ebf391e14388880c 127.0.0.1:7000 master,fail- 1478764594145 1478764593839 10 disconnected
　　79d8a5dc7e6140540c03fe0664df6a8150d0888f127.0.0.1:7004 slave 6d04c7a1efa63ca773114444591d04f6d340b07d 0 1478764605077 5connected
　　642a60187eea628a3be7fa1bb419e898c6b4f36a127.0.0.1:7002 master - 0 1478764604567 3 connected 10923-16383
　　bf2e5104f98e46132a34dd7b25655e41455757b5127.0.0.1:7005 slave 642a60187eea628a3be7fa1bb419e898c6b4f36a 0 1478764603546 6connected
　　7000日志：
　　32553:M10 Nov 16:02:11.016 * The server is now ready to accept connections on port7000
　　32553:M10 Nov 16:02:11.032 # Configuration change detected. Reconfiguring myself as areplica of c500c301825ceeed98b7bb3ef4f48c52d42fe72d
　　32553:S10 Nov 16:02:11.033 # Cluster state changed: ok
　　32553:S10 Nov 16:02:12.042 * Connecting to MASTER 127.0.0.1:7003
　　32553:S10 Nov 16:02:12.045 * MASTER  SLAVE sync started
　　32553:S10 Nov 16:02:12.045 * Non blocking connect for SYNC fired the event.
　　32553:S10 Nov 16:02:12.046 * Master replied to PING, replication can continue...
　　32553:S10 Nov 16:02:12.047 * Partial resynchronization not possible (no cached master)
　　32553:S10 Nov 16:02:12.049 * Full resync from master:08e765d07ea4bf0b0ce3f4e70a913047a9572ecd:1
　　32553:S10 Nov 16:02:12.146 * MASTER  SLAVE sync: receiving 18 bytes frommaster
　　32553:S10 Nov 16:02:12.146 * MASTER  SLAVE sync: Flushing old data
　　32553:S10 Nov 16:02:12.146 * MASTER  SLAVE sync: Loading DB in memory
　　32553:S10 Nov 16:02:12.146 * MASTER  SLAVE sync: Finished with success
　　32553:S10 Nov 16:02:12.173 * Background append only file rewriting started by pid32557
　　32553:S10 Nov 16:02:12.232 * AOF rewrite child asks to stop sending diffs.
　　7000端口恢复以后：
　　127.0.0.1:7001>cluster nodes
　　c500c301825ceeed98b7bb3ef4f48c52d42fe72d 127.0.0.1:7003 master - 01478765008832 11 connected 0-5460
　　6d04c7a1efa63ca773114444591d04f6d340b07d127.0.0.1:7001 myself,master - 0 0 2 connected 5461-10922
　　b62852c2ced6ba78e208b5c1ebf391e14388880c 127.0.0.1:7000 slavec500c301825ceeed98b7bb3ef4f48c52d42fe72d 0 1478765008328 11 connected
　　79d8a5dc7e6140540c03fe0664df6a8150d0888f127.0.0.1:7004 slave 6d04c7a1efa63ca773114444591d04f6d340b07d 0 1478765009839 5connected
　　642a60187eea628a3be7fa1bb419e898c6b4f36a127.0.0.1:7002 master - 0 1478765009839 3 connected 10923-16383
　　bf2e5104f98e46132a34dd7b25655e41455757b5127.0.0.1:7005 slave 642a60187eea628a3be7fa1bb419e898c6b4f36a 0 1478765009335 6connected
　　2.添加node节点：
　　首先，和上面一样加入7006,7007redis服务开启；
　　过程中7005找不到了，正好添加node就开始添加7005节点：
　　./redis-trib.rbadd-node --slave --master-id 642a60187eea628a3be7fa1bb419e898c6b4f36a127.0.0.1:7005 127.0.0.1:7003
　　然而报错了：
　　OK] All16384 slots covered.
　　Connectingto node 127.0.0.1:7005: OK
　　[ERR]Node 127.0.0.1:7005 is not empty. Either the node already knows other nodes(check with CLUSTER NODES) or contains some key in database 0.
　　在删除7005配置：rm -rf dump.rdb appendonly.aof nodes.conf以后还是报错，还好测试机上面没数据直接清库了……
　　root@localhostsrc]# redis-cli  -c -p 7005
　　127.0.0.1:7005>flushdb
　　root@localhostsrc]# ./redis-trib.rb add-node --slave --master-id642a60187eea628a3be7fa1bb419e898c6b4f36a 127.0.0.1:7005 127.0.0.1:7003
　　>>>Adding node 127.0.0.1:7005 to cluster 127.0.0.1:7003
　　[OK] Allnodes agree about slots configuration.
　　>>>Check for open slots...
　　>>>Check slots coverage...
　　[OK] All16384 slots covered.
　　Connectingto node 127.0.0.1:7005: OK
　　>>>Send CLUSTER MEET to node 127.0.0.1:7005 to make it join the cluster.
　　Waitingfor the cluster to join.
　　>>> Configure node as replica of 127.0.0.1:7002.
　　[OK] Newnode added correctly.
　　但是明明绑到7003上面，为什么现在是7002的从呢？
　　>>>Performing Cluster Check (using node 127.0.0.1:7005)
　　S:576437568beee83b412908e752a46a1273a076c5 127.0.0.1:7005
　　slots: (0 slots) slave
　　replicates642a60187eea628a3be7fa1bb419e898c6b4f36a
　　M:6d04c7a1efa63ca773114444591d04f6d340b07d 127.0.0.1:7001
　　slots:6461-10922 (4462 slots) master
　　1 additional replica(s)
　　M: 642a60187eea628a3be7fa1bb419e898c6b4f36a 127.0.0.1:7002
　　slots:0-999,5461-6460,10923-16383 (7461slots) master
　　2 additional replica(s)
　　S:b62852c2ced6ba78e208b5c1ebf391e14388880c 127.0.0.1:7000
　　slots: (0 slots) slave
　　replicates c500c301825ceeed98b7bb3ef4f48c52d42fe72d
　　S:bf2e5104f98e46132a34dd7b25655e41455757b5 127.0.0.1:7007
　　slots: (0 slots) slave
　　replicates642a60187eea628a3be7fa1bb419e898c6b4f36a
　　M: c500c301825ceeed98b7bb3ef4f48c52d42fe72d 127.0.0.1:7003
　　slots:1000-5460 (4461 slots) master
　　1 additional replica(s)
　　S:79d8a5dc7e6140540c03fe0664df6a8150d0888f 127.0.0.1:7004
　　slots: (0 slots) slave
　　replicates6d04c7a1efa63ca773114444591d04f6d340b07d
　　[OK] Allnodes agree about slots configuration.
　　>>>Check for open slots...
　　>>>Check slots coverage...
　　[OK] All16384 slots covered.
　　再检查下：./redis-trib.rb check 127.0.0.1:7005
　　Connectingto node 127.0.0.1:7005: OK
　　Connectingto node 127.0.0.1:7001: OK
　　Connectingto node 127.0.0.1:7002: OK
　　Connectingto node 127.0.0.1:7000: OK
　　Connectingto node 127.0.0.1:7003: OK
　　Connectingto node 127.0.0.1:7004: OK
　　>>>Performing Cluster Check (using node 127.0.0.1:7005)
　　S:576437568beee83b412908e752a46a1273a076c5 127.0.0.1:7005
　　slots: (0 slots) slave
　　replicates642a60187eea628a3be7fa1bb419e898c6b4f36a
　　M:6d04c7a1efa63ca773114444591d04f6d340b07d 127.0.0.1:7001
　　slots:6461-10922 (4462 slots) master
　　1 additional replica(s)
　　M:642a60187eea628a3be7fa1bb419e898c6b4f36a 127.0.0.1:7002
　　slots:0-999,5461-6460,10923-16383 (7461slots) master
　　1 additional replica(s)
　　S:b62852c2ced6ba78e208b5c1ebf391e14388880c 127.0.0.1:7000
　　slots: (0 slots) slave
　　replicatesc500c301825ceeed98b7bb3ef4f48c52d42fe72d
　　M:c500c301825ceeed98b7bb3ef4f48c52d42fe72d 127.0.0.1:7003
　　slots:1000-5460 (4461 slots) master
　　1 additional replica(s)
　　S:79d8a5dc7e6140540c03fe0664df6a8150d0888f 127.0.0.1:7004
　　slots: (0 slots) slave
　　replicates6d04c7a1efa63ca773114444591d04f6d340b07d
　　[OK] Allnodes agree about slots configuration.
　　OK,干净了。
　　接下来添加7006,7007两个节点，7006做主节点，7007做从节点：
　　[root@localhostsrc]# ./redis-trib.rb add-node 127.0.0.1:7006 127.0.0.1:7000
　　>>>Adding node 127.0.0.1:7006 to cluster 127.0.0.1:7000
　　Connectingto node 127.0.0.1:7000: OK
　　Connectingto node 127.0.0.1:7007: OK
　　Connectingto node 127.0.0.1:7001: OK
　　Connectingto node 127.0.0.1:7004: OK
　　Connectingto node 127.0.0.1:7002: OK
　　Connectingto node 127.0.0.1:7003: OK
　　Connectingto node 127.0.0.1:7005: OK
　　>>>Performing Cluster Check (using node 127.0.0.1:7000)
　　S:b62852c2ced6ba78e208b5c1ebf391e14388880c 127.0.0.1:7000
　　slots: (0 slots) slave
　　replicatesc500c301825ceeed98b7bb3ef4f48c52d42fe72d
　　M:6d04c7a1efa63ca773114444591d04f6d340b07d 127.0.0.1:7001
　　slots:6461-10922 (4462 slots) master
　　1 additional replica(s)
　　S:79d8a5dc7e6140540c03fe0664df6a8150d0888f 127.0.0.1:7004
　　slots: (0 slots) slave
　　replicates6d04c7a1efa63ca773114444591d04f6d340b07d
　　M:642a60187eea628a3be7fa1bb419e898c6b4f36a 127.0.0.1:7002
　　slots:0-999,5461-6460,10923-16383 (7461slots) master
　　2 additional replica(s)
　　M:c500c301825ceeed98b7bb3ef4f48c52d42fe72d 127.0.0.1:7003
　　slots:1000-5460 (4461 slots) master
　　1 additional replica(s)
　　S:576437568beee83b412908e752a46a1273a076c5 127.0.0.1:7005
　　slots: (0 slots) slave
　　replicates642a60187eea628a3be7fa1bb419e898c6b4f36a
　　[OK] Allnodes agree about slots configuration.
　　>>>Check for open slots...
　　>>>Check slots coverage...
　　[OK] All16384 slots covered.
　　Connectingto node 127.0.0.1:7006: OK
　　>>>Send CLUSTER MEET to node 127.0.0.1:7006 to make it join the cluster.
　　[OK] Newnode added correctly.
　　添加7007作为7006从节点：
　　./redis-trib.rb  add-node--slave --master-id 4d594ec18d8584b9d530d894f76098a6d4bd742a 127.0.0.1:7007127.0.0.1:7006
　　-slave从节点，masterid 主节点id ，127.0.0.1:7007 新节点，127.0.0.1:7006任意一个旧节点；
　　>>>Adding node 127.0.0.1:7007 to cluster 127.0.0.1:7006
　　Connectingto node 127.0.0.1:7006: OK
　　Connectingto node 127.0.0.1:7001: OK
　　Connectingto node 127.0.0.1:7000: OK
　　Connectingto node 127.0.0.1:7004: OK
　　Connectingto node 127.0.0.1:7002: OK
　　Connectingto node 127.0.0.1:7003: OK
　　Connectingto node 127.0.0.1:7005: OK
　　>>>Performing Cluster Check (using node 127.0.0.1:7006)
　　M:4d594ec18d8584b9d530d894f76098a6d4bd742a 127.0.0.1:7006
　　slots: (0 slots) master
　　0 additional replica(s)
　　M:6d04c7a1efa63ca773114444591d04f6d340b07d 127.0.0.1:7001
　　slots:6461-10922 (4462 slots) master
　　1 additional replica(s)
　　S:b62852c2ced6ba78e208b5c1ebf391e14388880c 127.0.0.1:7000
　　slots: (0 slots) slave
　　replicatesc500c301825ceeed98b7bb3ef4f48c52d42fe72d
　　S:79d8a5dc7e6140540c03fe0664df6a8150d0888f 127.0.0.1:7004
　　slots: (0 slots) slave
　　replicates6d04c7a1efa63ca773114444591d04f6d340b07d
　　M:642a60187eea628a3be7fa1bb419e898c6b4f36a 127.0.0.1:7002
　　slots:0-999,5461-6460,10923-16383 (7461slots) master
　　1 additional replica(s)
　　M:c500c301825ceeed98b7bb3ef4f48c52d42fe72d 127.0.0.1:7003
　　slots:1000-5460 (4461 slots) master
　　1 additional replica(s)
　　S:576437568beee83b412908e752a46a1273a076c5 127.0.0.1:7005
　　slots: (0 slots) slave
　　replicates642a60187eea628a3be7fa1bb419e898c6b4f36a
　　[OK] Allnodes agree about slots configuration.
　　>>>Check for open slots...
　　>>>Check slots coverage...
　　[OK] All16384 slots covered.
　　redis-cli-c -p 7006
　　127.0.0.1:7006>cluster nodes
　　4d594ec18d8584b9d530d894f76098a6d4bd742a 127.0.0.1:7006 myself,master - 0 0 13connected
　　6d04c7a1efa63ca773114444591d04f6d340b07d127.0.0.1:7001 master - 0 1478769599187 2 connected 6461-10922
　　b62852c2ced6ba78e208b5c1ebf391e14388880c127.0.0.1:7000 slave c500c301825ceeed98b7bb3ef4f48c52d42fe72d 0 147876960071011 connected
　　c500c301825ceeed98b7bb3ef4f48c52d42fe72d127.0.0.1:7003 master - 0 1478769601215 11 connected 1000-5460
　　7705e1148cdea9e5e1b05f4dab9a936e75857105127.0.0.1:7007 slave 4d594ec18d8584b9d530d894f76098a6d4bd742a0 1478769600205 13 connected
　　79d8a5dc7e6140540c03fe0664df6a8150d0888f127.0.0.1:7004 slave 6d04c7a1efa63ca773114444591d04f6d340b07d 0 1478769599187 2connected
　　642a60187eea628a3be7fa1bb419e898c6b4f36a127.0.0.1:7002 master - 0 1478769599700 12 connected 0-999 5461-646010923-16383
　　bf2e5104f98e46132a34dd7b25655e41455757b5:0 slave,fail,noaddr 642a60187eea628a3be7fa1bb419e898c6b4f36a 14787694043161478769403509 12 disconnected
　　576437568beee83b412908e752a46a1273a076c5127.0.0.1:7005 slave 642a60187eea628a3be7fa1bb419e898c6b4f36a 0 147876959970012 connected
　　节点看起来正常：但是bf2e5104f98e46132a34dd7b25655e41455757b5 :0slave,fail,noaddr 642a60187eea628a3be7fa1bb419e898c6b4f36a 14787694043161478769403509 12 disconnected
　　这个可能是在以前分配节点删除配置以后留下的，怎么处理？
　　重新分配solt：
　　[root@localhost src]#redis-trib.rb reshard 127.0.0.1:7001//下面是主要过程
　　How many slots do you want to move (from1to16384)? 1000//设置slot数1000

　　What is the receiving node>
　　Please enter all thesource node>　　Type'all'to use all the nodes as source nodes for the hash slots.

　　Type'done' once you entered all the source nodes>　　Source node #1:all //表示全部节点重新洗牌
　　Do you want to proceed with the proposed reshardplan (yes/no)? yes //确认重新分
　　[root@localhostsrc]# ./redis-trib.rb  reshard127.0.0.1:7006
　　Connectingto node 127.0.0.1:7006: OK
　　Connectingto node 127.0.0.1:7001: OK
　　Connectingto node 127.0.0.1:7004: OK
　　Connectingto node 127.0.0.1:7003: OK
　　Connectingto node 127.0.0.1:7002: OK
　　Connectingto node 127.0.0.1:7000: OK
　　>>>Performing Cluster Check (using node 127.0.0.1:7006)
　　S:bf2e5104f98e46132a34dd7b25655e41455757b5 127.0.0.1:7006
　　slots: (0 slots) slave
　　replicates642a60187eea628a3be7fa1bb419e898c6b4f36a
　　M:6d04c7a1efa63ca773114444591d04f6d340b07d 127.0.0.1:7001
　　slots:5461-10922 (5462 slots) master
　　1 additional replica(s)
　　S:79d8a5dc7e6140540c03fe0664df6a8150d0888f 127.0.0.1:7004
　　slots: (0 slots) slave
　　replicates6d04c7a1efa63ca773114444591d04f6d340b07d
　　M:c500c301825ceeed98b7bb3ef4f48c52d42fe72d 127.0.0.1:7003
　　slots:0-5460 (5461 slots) master
　　1 additional replica(s)
　　M:642a60187eea628a3be7fa1bb419e898c6b4f36a 127.0.0.1:7002
　　slots:10923-16383 (5461 slots) master
　　1 additional replica(s)
　　S:b62852c2ced6ba78e208b5c1ebf391e14388880c 127.0.0.1:7000
　　slots: (0 slots) slave
　　replicatesc500c301825ceeed98b7bb3ef4f48c52d42fe72d
　　[OK] Allnodes agree about slots configuration.
　　>>>Check for open slots...
　　>>>Check slots coverage...
　　[OK] All16384 slots covered.
　　How manyslots do you want to move (from 1 to 16384)?1000
　　redis-trib.rb reshard 127.0.0.1:7006
　　根据提示选择要迁移的slot数量
　　(ps:这里选择1000)
　　How many slots do you want to move (from 1 to 16384)? 1000
　　选择要接受的这些soltnode_id
　　What is the receiving node ID? f51e26b5d5ff74f85341f06f28f125b7254e61bf
　　选择slot来源:   #all表示从所有的master重新分配，
　　#或者数据要提取slot的master节点id,最后用done结束

　　Pleaseenter all the source node>　　Type 'all'to use all the nodes as source nodes for the hash slots.

　　Type'done' once you entered all the source nodes>　　Sourcenode #1:all
　　#打印被移动的slot后，输入yes开始移动slot以及对应的数据.
　　#Do youwant to proceed with the proposed reshard plan (yes/no)? yes
　　#结束
　　但是现在的疑问，迁移的slot数量怎么确定？
　　查看nodes：
　　redis-cli  -c -p 7007
　　127.0.0.1:7007>cluster nodes
　　c500c301825ceeed98b7bb3ef4f48c52d42fe72d127.0.0.1:7003 master - 0 1478770184173 11 connected 1272-5460
　　6d04c7a1efa63ca773114444591d04f6d340b07d127.0.0.1:7001 master - 0 1478770182154 2 connected 6733-10922
　　642a60187eea628a3be7fa1bb419e898c6b4f36a127.0.0.1:7002 master - 0 1478770183165 12 connected 456-999 5461-646010923-16383
　　4d594ec18d8584b9d530d894f76098a6d4bd742a 127.0.0.1:7006 master - 01478770183165 13 connected 0-455 1000-1271 6461-6732
　　7705e1148cdea9e5e1b05f4dab9a936e75857105 127.0.0.1:7007 myself,slave4d594ec18d8584b9d530d894f76098a6d4bd742a 0 0 0 connected
　　b62852c2ced6ba78e208b5c1ebf391e14388880c127.0.0.1:7000 slave c500c301825ceeed98b7bb3ef4f48c52d42fe72d 0 147877018417311 connected
　　79d8a5dc7e6140540c03fe0664df6a8150d0888f127.0.0.1:7004 slave 6d04c7a1efa63ca773114444591d04f6d340b07d 0 1478770184174 2connected
　　576437568beee83b412908e752a46a1273a076c5127.0.0.1:7005 slave 642a60187eea628a3be7fa1bb419e898c6b4f36a 0 147877018265812 connected
　　以上分的好碎，新添加的主节点是没有slot的,主节点如果没有slots的话，存取数据就都不会被选中。所以重新洗牌后就新加的master node 就有slot了。 注意：如果用’all’则会将其他每个master节点的slot 分给目标节点。最好别用‘all’， 会造成slot 分割碎片。
　　3.删除节点：
　　3.1 删除从节点一条命令就可以：
　　./redis-trib.rbdel-node 127.0.0.1:7007 'bf2e5104f98e46132a34dd7b25655e41455757b5'
　　>>>Removing node bf2e5104f98e46132a34dd7b25655e41455757b5 from cluster127.0.0.1:7007
　　Connectingto node 127.0.0.1:7007: OK
　　Connectingto node 127.0.0.1:7005: OK
　　Connectingto node 127.0.0.1:7001: OK
　　Connectingto node 127.0.0.1:7003: OK
　　Connectingto node 127.0.0.1:7002: OK
　　Connectingto node 127.0.0.1:7004: OK
　　Connectingto node 127.0.0.1:7000: OK
　　>>>Sending CLUSTER FORGET messages to the cluster...
　　>>>SHUTDOWN the node.
　　3.2 删除主节点：
　　注意：如果主节点有从节点，将从节点转移到其他主节点在删除；
　　127.0.0.1:7007> cluster repicate642a60187eea628a3be7fa1bb419e898c6b4f36a
　　OK
　　642a60187eea628a3be7fa1bb419e898c6b4f36a是7002节点id；
　　# redis-trib.rb reshard 127.0.0.1:7006 //取消分配的slot,下面是主要过程
　　How manyslots do you want to move (from1to16384)? 1000 //被删除master的所有slot数量

　　What is the receiving node>
　　Please enterall the source node>　　Type 'all'to use all the nodes as source nodes forthe hashslots.

　　Type 'done' once youentered all the sourcenodes>　　Source node #1:4d594ec18d8584b9d530d894f76098a6d4bd742a //被删除master的node-id
　　Source node #2:done
　　Do you want to proceed withthe proposedreshard plan (yes/no)? yes //取消slot后，reshard
　　注意：source node 要用 ‘done’,如果用all 则会全部洗牌，7006还是会得到slot.
　　最后删除主节点7006:
　　./redis-trib.rbdel-node 127.0.0.1:7006 '4d594ec18d8584b9d530d894f76098a6d4bd742a'
　　./redis-trib.rbdel-node 127.0.0.1:7006 '4d594ec18d8584b9d530d894f76098a6d4bd742a'
　　>>>Removing node 4d594ec18d8584b9d530d894f76098a6d4bd742a from cluster127.0.0.1:7006
　　Connectingto node 127.0.0.1:7006: OK
　　Connectingto node 127.0.0.1:7001: OK
　　Connectingto node 127.0.0.1:7000: OK
　　Connectingto node 127.0.0.1:7003: OK
　　Connectingto node 127.0.0.1:7007: OK
　　Connectingto node 127.0.0.1:7004: OK
　　Connectingto node 127.0.0.1:7002: OK
　　Connectingto node 127.0.0.1:7005: OK
　　>>>Sending CLUSTER FORGET messages to the cluster...
　　>>>SHUTDOWN the node.
　　查看结果：
　　redis-cli  -c -p 7000
　　127.0.0.1:7000>cluster nodes
　　6d04c7a1efa63ca773114444591d04f6d340b07d127.0.0.1:7001 master - 0 1478771310480 2 connected 6733-10922
　　bf2e5104f98e46132a34dd7b25655e41455757b5 :0 slave,fail,noaddr642a60187eea628a3be7fa1bb419e898c6b4f36a 1478769404363 1478769403658 12disconnected
　　79d8a5dc7e6140540c03fe0664df6a8150d0888f127.0.0.1:7004 slave 6d04c7a1efa63ca773114444591d04f6d340b07d 0 1478771309976 5connected
　　b62852c2ced6ba78e208b5c1ebf391e14388880c127.0.0.1:7000 myself,slave c500c301825ceeed98b7bb3ef4f48c52d42fe72d 0 0 10connected
　　c500c301825ceeed98b7bb3ef4f48c52d42fe72d127.0.0.1:7003 master - 0 1478771308967 11 connected 1272-5460
　　642a60187eea628a3be7fa1bb419e898c6b4f36a127.0.0.1:7002 master - 0 1478771309976 14 connected 0-1271 5461-673210923-16383
　　576437568beee83b412908e752a46a1273a076c5127.0.0.1:7005 slave 642a60187eea628a3be7fa1bb419e898c6b4f36a 0 147877130896714 connected
　　这样去掉了7006,7007，redis块还是0-1271，1272-5460，5461-6732，6733-10922，10923-16383但是红的部分是什么呢？待查明！